摘要
Theobjectiveofthisstudywastoobtainspatialdistributionmapsofpaddyricefieldsusingmulti-datemoderate-resolutionimagingspectroradiometer(MODIS)datainChina.Paddyricefieldswereextractedbyidentifyingtheuniquechar-acteristicofhighsoilmoistureinthefloodingandtransplantingperiodwithimprovedalgorithmsbasedonricegrowthcalendarregionalization.Thecharacteristiccouldbereflectedbytheenhancedvegetationindex(EVI)andthelandsurfacewaterindex(LSWI)derivedfromMODISsensordata.Algorithmsforsingle,early,andlatericeidentificationwereobtainedfromselectedtypicaltestsites.Thealgorithmscouldnotonlyseparateearlyriceandlatericeplantedinthesamefields,butalsoreducetheuncertainties.ThearealaccuracyoftheMODIS-derivedresultswasvalidatedbycomparisonwithagriculturalstatistics,andthespatialmatchingwasexaminedbyETM+(enhancedthematicmapperplus)imagesinatestregion.Majorfactorsthatmightcauseerrors,suchasthecoarsespatialresolutionandnoisesintheMODISdata,werediscussed.Althoughnotsuitableformonitoringtheinter-annualvariationsduetosomeinevitablefactors,theMODIS-derivedresultswereusefulforobtainingspatialdistributionmapsofpaddyriceonalargescale,andtheymightprovidereferenceforfurtherstudies.
出版日期
2009年10月20日（中国期刊网平台首次上网日期，不代表论文的发表时间）
